Project: replacing KS18XL inner shell

[svenomous]

My KS18XL needs a new inner shell. The old shell has cracked badly at the 12 machine screws that hold the case to the pedal hangers. The problem started last year (June) when I got the new wheel and the tire was rubbing the case on one side. I found the wheel to be off-center, so I opened up the case to wiggle things around and get it centered. I had trouble with the Phillips screw heads (my screwdriver was stripping the heads) and changed them to hex bolts, and then I over-torqued the new bolts. I put in some plastic washers to distribute the pressure better, and added hot glue, but overall it was a pretty ugly situation. I changed out the bolts and put in Torx-driven screws instead, but the damage was done. I've ridden the wheel since then (with a long break for fall/winter/spring), but in the last 2 weeks it started creaking a lot. The case being secure to the actual wheel/motor/pedals assembly is very important, since the case is where the mainboard is, and the accelerometers that keep the whole thing balanced. A case that's loose would cause very strange oscillations I'm sure.

So, I ordered a new shell (both sides) from Jason at eWheels. Ordered the white shell for my black wheel, because apparently the white plastic is known to be less brittle, and because Jason assured me that the side panels, front and back panels, and top panel completely cover the inner shell and no part of the white will show through.

The shell was scheduled to be delivered today (Saturday), and it's raining, so I got started on the project this afternoon. I've finished the disassembly (2.5 hours) and will report on that in this post, with the assembly following in another post. FedEx has since updated the delivery to tomorrow (Sunday), so hopefully it will arrive early enough in the day and I will be able to finish the project this weekend

Removing the LED strips and side panels was simple, as was disconnecting the motor and battery cables and removing the 12 screws holding the case to the pedal hangers. See what the plastic looked like at the shell/hanger connection points (pretty bad, that's what it looked like):

 (screws still on, with washers, left side of wheel)

 (screws off, left side of wheel)

 (screws off, right side of wheel)

Tip: remove the side panel pads very carefully (so they don't rip), and stick them to a piece of wax paper so their adhesive backings don't get too degraded by dust and debris:

Tip: make notes and take photos for reference. I made notes about the speaker orientations, which side of the pedals the threads are on (the large KS18XL pedals have just one retaining screw, on one side of the pin, pedal retaining pin orientation (it has a rounded edge and a flat edge, orientation of the motor (which side the main power cable is on, vs. the hall sensor cable. In addition to the photos for this show-and-tell post, I also took photos of how some of the wiring was routed in the case, and of how the cables are bundled (with zip ties) and routed at the mainboard.

While I was at it I removed the pedals. This isn't necessary for the procedure (the pedals can just stay attached to the pedal hangers), but I wanted to clean and inspect for damage (cracks). The pedals are pretty scuffed from past falls and scrapes, but they look structurally sound:

As soon as I pulled the motor/wheel assembly out of the case, that part of the case showed just how badly it was broken. No wonder it was starting to creak, and I'm sure it was not far from completely cracking free and starting to move around against the wheel assembly:

Tip: be careful while removing the motor/wheel assembly from the case. It slides out, and there are wires that are threaded through holes in the pedal hangers to enter the case; these wires need to pull out with the rest of the assembly, and it's easy for them to get caught up or pinched, and to end up with damaged insulation or worse. So, be slow and deliberate, and watch that the cables are feeding through properly while sliding out the assembly. When the whole thing is out, inspect the wires for any damage. Be equally careful and inspect closely when sliding the thing back together and threading those wires back through the hanger holes.

With the motor/rim/tire/tube/pedal-hanger assembly out of the case, and the two battery connectors disconnected from the mainboard, it was time to drain the capacitors, so I pushed in and held the power button, got about 3 seconds of very loud beeping and a click or two, and let go once the board seemed dead.

Next came the battery packs. These have double-sided tape on the bottom to hold them in place, as well as some foam pads at the bottoms to cushion them and hold them. I carefully pried the larger pad on one side loose and removed it (careful, it's lightly glued to the case, so it can rip, so it's a good idea to use a flat-head screwdriver to carefully scrape along the plastic to detach it). I also stuck the flat-head between the pack and the other two bottom pads, loosening the adhesive bonds there. With room opened up to get my fingers under the pack, I reached in and very slowly and carefully pulled up from the bottom corner, letting it detach little by little under pressure and getting more and more of my fingers and hand (and a long screwdriver) underneath the pack, trying to support as much of the pack as possible to prevent too much torsion on the fragile wrapping. Pulling too fast would cause the adhesive tape to rip and stay behind on the case, and could also rip the battery pack wrapping. Ripping the wrapping would cause the whole pack to flex, which could damage connections between cells, and between the cell groups and the BMS. Got both packs and their pads removed cleanly, and set them aside.

Next, more screws: those holding the front/rear trim in place, those holding the two shell halves together, those holding the trolley handle in place, those attaching the speakers and buzzer to the case, and a couple holding wire clips in place. Tip: There are many screws of various types (machine screws, plastic screws), lengths, etc., so another tip: put the screws in separate cups or piles or something, maybe with a label or two to identify where they go:

There are some foam blocks with light adhesive tape on them, holding down wiring in each case half. I removed these carefully and set them aside, and took a couple photos to remember where they go.

I carefully pried up the top trim/cover (which covers the mainboard) and loosened the front and rear trim pieces. Raised the trolley handle all the way up and raised the cover up along the trolley handle rods to get it out of the way. Carefully detached the bottoms of the trolley handle rods from the case, with lift sensors still attached to them (I saw no need to remove the sensors), and pulled the whole trolley handle assembly, with top cover, off the case:

There are quite a few wires coming down from the mainboard and going to each side of the case, as well as to the front and rear panels (charge ports, USB ports, power switch, head/tail lights). Some connectors can only go to one thing, but not all of them (there were a couple that were the same). Also, it's easy to get left/right confused. So, another tip: label the wiring, marking the ends as going to "left front," or "front panel," or "rear left" for example; I used the power button as my "front" reference, and labeled everything in relation to that. Here you can see some of my labeled wires (just scotch tape with some marker-scribbling on it):

There are quite a few zip ties bundling wires neatly and helping to route them. A couple of the zip ties are threaded through holes in the side of the mainboard as well, which keeps those wire bundles in position. Tip: for reassembly purposes in the new case, acquire a bunch of small zip ties. I have an assortment that includes smaller/thinner ones:

Tip: get some grease for the reassembly. The retaining pins should probably be lightly greased during reassembly, which will help prevent or reduce corrosion, and will also make the pedals fold and unfold a more smoothly.

Before long I had all the connectors on the mainboard removed (with one exception they're unique, and the two that are identical I labeled as "front" and "back"), all zip ties cut off, individual wire pigtails removed, the headlight/taillight removed, and the front and rear panels removed. The only wire that doesn't connect to the mainboard at one end is the small thin brown single wire with a two-pin JST connector at each end that connects to each battery pack; this one just crosses the mainboard from side to side (neatly bundled to the left and right side of the mainboard via zip ties), to reach the area where each battery pack 's main power connector gets connected to the mainboard. The mainboard itself unscrews from the case with 4 screws. Lifted off the board and set it aside. That's it!  At this point the case had nothing left on it except the gasket that seems to protect the mainboard from water/dust from the wheel: the mainboard sits right on top of the wheel well, with a heat sink sticking through a rectangular opening into the wheel well. I guess it gets cooling airflow from the rotating wheel during operation. The aluminum panel and heat sink were dusty but not too bad, considering they're directly facing the tire at the top of the wheel well. The gasket I left on the old case, as it's very thin and fragile, and I don't know if the new shell will come with its own. If it doesn't, I'll have to try to transfer the existing gasket to the new shell.

All that was left was to clean all the parts. There were a lot of plastic bits and shavings from the broken case to vacuum up, and I blew some air on the mainboard, wiped the heatsink/panel on the bottom of the board, and wiped down the motor/wheel/rim and side panels. This was a good opportunity to get a close look at the tire and wheel, to make sure there's no hidden damage. Here are some more photos of the battery packs, the removed parts, bare inner case, cleaned panels, and cleaned wheel assembly:

Edited May 17, 2020 by svenomous
Removed mention of thread-locker. See my later post in this thread about how that stuff is like acid to the inner shell pastic!

[svenomous]

By the way, PSA: do safety checks every once in a while: pedals can crack and suddenly break under load when you least expect it; check that screws are tight. I forgot to mention above that when I removed the pedals, the set screws were very easy to remove, they were not tight at all, and the little spring plates that are attached via two screws were half loose, both of them.  On the other hand, I found that all the bolts holding the motor case to the wheel, and the extremely important large bolts fixing the pedal hangers to the axle were nice and secure. The plugs were secured with zip ties, which is a good idea, but I had forgotten to file the edges round on the zip ties, so when I removed them today I found insulation on some wires to have been slightly abraded; not quite damaged, but enough vibration could have caused a zip tie edge to saw through to the wire eventually. The live 84V wire that my safety depends on. So: occasional safety inspection  strongly recommended. Also check tire (is it "true", is the rim bent or dinged anywhere, are there any thumbtacks in it?), maybe take the motor covers off and check the axle.  I haven't done that last thing yet, but am considering it before assembly tomorrow.

[Preston Palmer]

Nice work and awesome write-up!  I have a black shell 18XL that did a pedal hanger crack repair on a few months ago with some epoxy and washers.  It stiffened it up for sure but I assume I will have to swap in a whole new case like you did soon enough.  As for the zip ties.  Consider the GripLockTies that are rubber lined.  They help with wire abrasion and vibration.  Aviation industry uses them,  https://www.amazon.com/GRIPLOCKTIES-Releasable-Re-Usable-Industrial-Durable/dp/B07CZMQB4K

The 18XL is such as awesome wheel! Enjoy!

[svenomous]

Thanks @Preston Palmer, will check out those GripLockTies. For the moment, to complete the build, I'll use regular ties with edges a bit filed down to round them, and also with some electrical tape as a buffer. If/when I get something better, I can open up again and replace.

The shell did arrive this morning, and I've been working on it. There was a delay because I had to apply silicone sealant and wait for it to partially cure (more on that in the next write-up), and I'm also really taking my time to make sure I don't make any mistakes. Should have the reassembly write-up ready in a few hours.

[svenomous]

Well, there've been two hiccups.  I finished, but something's wrong, and something broke.  I'm starting two new threads to talk about these things. Won't post the "reassembly" post with photos until those issues are resolved.

[svenomous]

I had two issues on the rebuild. One was simple (just a reversed connector). The other was that I cracked the new shell, if you can believe it. See my other thread if interested. There is one important difference between my old black shell (1st gen?) and the new white shell I received, which combined with some frustration on my part caused me to crack my beautiful new shell that I'd had for only a few hours.  So sad/maddening/frustrating! I've got another thread going on what to do next, and it looks like I'll be attempting a repair, which should be possible with the white plastic vs. the old very brittle black shell.

Edit: I've tried to remove or annotate any mention of thread-locker in this forum thread. Don't use the stuff anywhere that might come in contact with the inner shell plastic.  It's what caused the cracking, as I later found out.

OK, so here's the narrative on the re-build, with photos. The important design difference that tripped me up is explained, as well as what you might have to do to avoid what happened to me. Enjoy.

New shell arrived this morning. More precisely, two new shell halves arrived. The halves are identical, and at first glance when you put them together the result is identical in shape and details, compared to the old one:

The only obvious difference is that the screw posts are designed slightly differently, as are the metal inserts of the posts under the curved LED strips. 8 screws bring the halves together., but I already removed those from the old shell, so what was still holding it together in one piece? Took a closer look inside the older shell and saw:

Some kind of adhesive was painted on the inside, probably also on the edges before the halves were joined. I don't know what the stuff is, exactly, but I decided I'd use silicone sealant:

The stuff takes a pretty long time to set, but it starts getting tacky within a few minutes, so I worked fast, using my finger, along all the edges that press together:

Then I pressed the halves together and inserted/tightened the 8 screws that hold them together. Using more sealant and a finger, I worked around the outside and inside to fully seal. The sealant is not just a glue in this case, it also provides some of the water resistance of the final product.

The little rectangular foam gasket/strip that goes around the edge of the opening the mainboard sticks its heat sink through was not included with the new halves, so I carefully pried it up with a flat-head screwdriver, found that its adhesive was still good, and applied it to the new shell:

As I tried to screw the mainboard onto the shell, I noticed that there was nothing to screw it to. Looked at the old shell and saw that there were 4 small nuts inserted into the plastic and held in place with silicone sealant:

Pried them out, removed the sealant, and inserted them into the new shell. Aligned them, screwed in the mainboard, and applied silicone sealant:

The mainboard in place, it was time to deal with the front and rear panel., and the headlights. First inserted the headlights:

As I started inserting connectors, I also started putting dabs of silicone on the edges to help keep them in place:

I've seen questions on the forums about charge current for the KS18L/XL, so since I had it all apart I though I'd show you what's going on at the dual charge port:

As you can see, you have a mini circuit board that parallels the two connectors, and a single pigtail comes out of it and connects to a pigtail coming out of the mainboard. Without cutting the wire, I couldn't tell the wire gauge, but based on total wire thickness with insulation it looks to be 14 gauge. Don't take that as gospel, though.

Next, connected, threaded wires, and inserted front and rear panel, remembering to plug in the headlamps while the connectors were accessible:

With mainboard in place, and headlamps and front/rear panel connected, I needed to route the remaining wires and hook up a couple of loose items (like the buzzer). First, I loosely arranged the wires to go to the correct quadrants (front right/left and rear right/left), which is when all the labels I put on the wires during disassembly came in very handy:

Started applying zip ties to bundle the wires, remembering that there are some holes in the mainboard for the zip ties to anchor to:

I had one moment of panic at this point. I'd attached everything to the mainboard and applied dabs of sealant, and I'd just finished zip-tying everything, when I noticed one socket with nothing plugged in to it. You can see it in the lower right quadrant of the previous photo. I couldn't find anything that should go there. What was happening? Where was it? What had I missed? So I went back to the photos taken during disassembly, and...the connector never had anything plugged into it.  It's supposed to be unpopulated. Phew!

Time to turn my attention to the left and right side. There are a bunch of little foam block, which I recovered from the old shell, which are used for cable management. They hold down wires and keep them from flopping around and making noise while riding. These have weak adhesive on them. I didn't have reposition-capable adhesive on hand, but I did have 3M VHB tape (a double-sided adhesive foam tape for mounting purposes), which I started applying in patches to the foam blocks to "refresh" their weakened adhesives. The VHB is a little too strong for this application, but it is removable, sort of, and it worked great for me. Here's a photo of the roll of tape and some of the foam blocks:

I also had to "fabricate" some replacement blocks that I was missing for some reason, for which I used some closed-cell white foam weatherstripping I scavenged from my household supplies. You can see a couple of these in the previous photo.

Mounted the 4 speakers and the buzzer (which goes on the right side, and connected the speakers:

Time to get the trolley handle installed. This is when I noticed the most obvious design change between the old case and the new one:

So on the old one, the bottom of the long slot the trolley rail goes into is open, which is why I was able to remove the whole trolley mechanism along with the attached sensor plates/screws. On the new one, there's some extra plastic and the adjustment screw is supposed to go through a hole in the plastic. So, simply inserting the rails wasn't an option:

I had to unscrew the sensor screw and plate, insert the rail, and screw the sensor mechanism back in. From what I read, there's no specific "adjustment" for the sensor, it just needs to be screwed in to remain loose enough so the little metal plate can still move around a little bit. So I screwed in the plate, left it loose enough to "wiggle a bit," and applied silicone sealant to fix the screw position, similar to how some kind of red goop was on the screw originally to fix it in place. Don't know if the sensor will work correctly this way, but I don't care since I keep the trolley sensor disabled anyway:

Trolley handle installed, with retaining metal strips in place:

OK, here's an important piece of information. On the old case, the metal strips that hold down the trolley handle rails is secured by plastic screws, and there are little plastic spacer washers underneath the strips. Here's a photo of a strip with the screws and washers:

However, when installing the strips on the new case, note how the metal doesn't touch the rail:

I knew something wasn't right, but stupidly I just accepted it and moved on. Later this mistake led to side covers that didn't fit (the screws stick out too much), and this caused me to push too hard to try to make the covers fit, and that caused me to crack the new case by flexing it. Duh. So, look closely and if the metal strip is hovering above the rail when you install it, remove the spacer washers. Clearly the screw posts for the trolley rails on the new white shell are taller and incorporate the extra height that was previously provided by the spacers. A design improvement I wish I'd known about ahead of time.

The battery packs are heavy, so I decided to install the wheel assembly first, since that requires some manhandling and wire threading. I placed the wheel assembly upright on the ground and lowered the shell down onto it after noting orientation (one side of the shell has the two big connectors , and the other side has one big connector and one smaller connector for the hall sensor...this has to line up with the correct sides of the motor housings). I aligned the big plastic slots with the pedal hangers, threaded the wires through the holes in the shell, and slowly lowered the shell onto the hangers while keeping the wires running through the raised channels in the plastic and preventing things from getting tangled up. Tip: wrap some electrical tape around each wire bundle coming out of the motor halves, especially the thick one with the main power wires, because the plastic shell has sharp edges and will easily abrade/cut the insulation on the wires. I didn't do this, and this is the result:

Didn't quite make it all the way through to bare wire, but close! Electrical tape to the rescue:

Everything beautifully aligned, and such beautiful uncracked plastic:

Used plastic washers to hopefully distribute the stress a bit. Didn't tighten screws yet, so just screwed in until they touch bottom:

Attached the foam pads that cushion the bottom of each battery pack (as noted above, I improved their adhesives first with some double-sided mounting tape):

Inserted each battery pack. The double-sided adhesive tape lining the bottom of each pack was still pretty strongly adhesive, and the packs seemed to adhere well enough to the case. Threaded the motor and battery wires through the paper sleeves:

Tip: before connecting the batteries to the mainboard and to each other, measure each one's voltage. Most importantly, doing this told me that each pack was at the same voltage. If the packs are at different voltages (not within half a volt or so), bad things could happen when they get paralleled through the mainboard, as large voltage differences cause arcing, and as the lower-voltage pack would start taking a high-amperage charge from the higher-voltage pack (the paralleled packs seek to achieve equilibrium at the same voltage). I found that both packs were at 81.7V. When making the first connection (connecting the first pack to the cable going to the mainboard), there will be a voltage differential as the capacitors take a charge. This can cause a spark (arcing), so I suggest you wear gloves or use a couple of shop towels, and that you put the connectors together quickly and firmly. Doing it slowly or hesitating/fumbling can make the spark worse. Prepare yourself (be mentally ready for a "pop" sound and a visible spark), get the connectors aligned, and drive them together. Usually when the second pack is connected, things are more tame, assuming the pack voltages are the same.  I still suggest protecting hands just in case.

I prefer to have the big connectors (the ones carrying motor power and battery power, so 3 of them for this wheel) secured with zip ties to prevent them from working loose due to vibration/shocks. Even thin zip ties like I'm using have sharp edges that can eventually saw through the insulation in wires on the connectors. I just learned from @Preston Palmer about rubber-cushioned zip ties (thanks!), but since I didn't have these yet, I applied electrical tape to the zip ties I used to secure the connectors:

Then I zipped the tie around each of the big connectors, threading it between wires and zipping it so the connectors cannot disconnect:

Connected the hall sensor wires and the little single-pin wires between the batteries (anyone know what these are for?), and applied the big foam blocks that hold down the battery/motor wiring:

OK, time to tighen down the shell and try to center it over the wheel. The next photo doesn't show alignment very well. On one side I got it perfect, but on the other there was a slight bias to one side that I couldn't get rid of. The screw holes don't allow for much play:

At this point I also reinstalled the pedals, obviously. I used blue thread-locker on the pin-retaining screws, and on the small screws holding the spring plates in plate, and I lightly greased the pins themselves with some general-purpose medium grease I use for my bike work.

Edit: I believe it's OK to use thread-locker on the pin-retaining screws, and on the screws holding the spring plates to the pedals.  These are metal parts screwing into metal.  Just avoid using the stuff if there's any chance it will touch the inner shell plastic, as that will cause the plastic to essentially come apart, as I found out the hard way.

That's it! All that was left (which I didn't bother photo-documenting) was to attach the side panels, threading the LED strip wires/connectors through them, screwing in the 15 screws on each side panel (15!), and connecting and snapping in the LED strips. One little SNAFU was that one of the LED strips was connected backward initially. This shouldn't be possible since the connectors are keyed, but the keyed outer sleeve actually pulled free on one of the wires, and I accidentally put it back on backward. The result was a wheel that was making intermittent clicking noises and flickering the headlights and LED strips, except the strip with the reversed connector. I figured it out, turned things the right way, and the wheel works great now. Haven't calibrated it yet, although it should definitely be calibrated after this type of operation. The accelerometers in the mainboard need to be told what "perfectly level" looks like in the new shell.

Haven't ridden yet, since (a) it was dark by the time I finished, and (b) I managed to crack the new shell, as previously explained, and I want to make sure it's safe to ride without making the cracks worse, while I figure out how to repair the cracks.

Edited May 17, 2020 by svenomous
Removed or annotated mentions of thread-locker in this post, to try to prevent anyone from making the mistakes I made.

[seage]

I did this swap over last year for the same reason, and it was NOT fun. I was annoyed that it cracked so easily in the first place but what can ya do. When i got the white case i was surprised to see that they made absolutely no improvements on the design or any reinforcements or anything at all after probably 8 billion ppl complaining LOL. 

We did it a little differently as we kept the motherboard fully plugged in, tore down the body and then put everything together around the motherboard again. I know for a fact, even with labels, my whole wheel would explode had i disconnected anything. 

[svenomous]

4 hours ago, seage said:

I did this swap over last year for the same reason, and it was NOT fun. I was annoyed that it cracked so easily in the first place but what can ya do. When i got the white case i was surprised to see that they made absolutely no improvements on the design or any reinforcements or anything at all after probably 8 billion ppl complaining LOL. 

We did it a little differently as we kept the motherboard fully plugged in, tore down the body and then put everything together around the motherboard again. I know for a fact, even with labels, my whole wheel would explode had i disconnected anything. 

If nothing else, the white plastic is apparently a lot less brittle. I can vouch for the fact that the black plastic on mine wasn't just cracking, it was turning into a fine powder. So the white shell represents a design improvement from a materials perspective, for someone with a black inner shell.

Haha, it never occurred to me to try to keep everything connected! How does that even work, since when you remove the front/rear panel it has to pull its pigtail wires away with it, and some of them attach at the motherboard end, not at the panel end?

[svenomous]

So...I'm having to re-do the whole thing. The reason is that the new shell has also cracked. For my second time around, the plan is to reconsider the nylon washers (not sure whether they're doing the good I intended them to), to tighten the bolts that hold the shell onto the pedal hangers very lightly, and to rely on the thread-locker to keep these very loose bolts from backing out later.

Edit: don't use thread-locker! See my post below. The thread-locker was what was cracking the shell, as I now know.

I'm not going to document the second time around. Disassembly went pretty fast without stopping to take photos (and with some experience under my belt), and I thought I'd report on one interesting thing I noted:

Use of silicone sealant to "glue" the two shell halves together and create a waterproof barrier against water/dust ingress from the wheel well was not exactly a failure, but it also wasn't a great success. If you've worked with this type of sealant before, it takes a couple of hours to set enough to not be very "tacky" anymore, and is supposed to be fully set/cured within 24 hours. However, it remains kind of soft and gummy for quite a while, weeks or months, before eventually settling into the hard rubbery consistency it has in its more mature existence. What this means in my case is that (a) the sealant spread along the inside seam in the wheel well permanently collected a an ugly layer of dust from 2 trail rides I went on; (b) the headlights were slightly glued in place by being in contact with the sealant that had squeezed out along the seam at the headlamp installation holes; (c) the mainboard's bottom metal plate had glued itself to the sealant at the seam, and when I removed the board a bunch of sealant came up with it, pulling right out of the gap between the shell halves as well.

I think this stuff does work, sort of, but it's just a little too soft for the intended purpose. I've ordered a flexible plastic bonding agent to try to use for the second rebuild. It apparently sets strongly, retains some flexibility, isn't tacky once set, and is not silicone, although it is waterproof. If that doesn't bond well, I still have plenty of silicone to use as a backup.

Edited May 17, 2020 by svenomous
Annotated a warning about thread-locker use.

[svenomous]

PSA (which I will repeat in its own thread, with an appropriate title): do not apply thread-locker when doing one of these builds, at least not with the KS inner shells.  See this thread:

In the end, I repeated the above-described project 3 times (!), because the first and second time I cracked the shell due to the use of thread-locker (Loc-Tite for example).  The stuff is like acid to this inner shell plastic. On the second build I watched it create cracks that weren't there before, right in front of my disbelieving and appalled eyes!

I've removed the reference/tip about using thread-locker in the above posts, to make sure nobody actually tries to use the stuff because of something I said in my write-ups.

[svenomous]

A follow-up:

Opened up the wheel today, after a few rides, to see if everything looks good, and it does!  No cracks. Lightly tightened a few of the pedal hanger attach bolts, since I had snugged them down very lightly and a couple felt a little loose.  

Two little items that might be useful to someone:

First: When adjusting the lift sensor, the screw has to be tightened to a certain point, slightly tight but not too tight. It's not really tight enough to ensure the screw doesn't back out eventually, especially given the vibrations that are part of operation. Originally the set screw had a slightly flexible red goop on it, which I didn't have, so I used silicone sealant to try to keep the screw from turning, but the stuff is too soft. Did some research, and found out what the red stuff is (or at least, what people often use for set screws like this): nail polish! So I picked up some good old regular red nail polish from the beauty aisle at my grocery store, and painted some on the screw edge (and sloppily all over the surrounding metal as well):

Second: If you remove the pads from the outer case frequently (to get at the screws they cover), the adhesive starts to weaken/fail. I usually put the pads on wax paper to keep dust from settling on the adhesive, but after repeated removal/re-installation they start to lift up at the edges. Well, I found something that "refreshes" the adhesive. It's used for home arts/crafts projects, e.g. to stick sequins onto fabric and suchlike.  Seems to work pretty well so far, see photos below.

It's a white liquid, which is applied and allowed to dry (it becomes transparent and retains a tacky feel). I applied liberally to each pad, waited for the stuff to dry, and applied the pads: they adhere to the case pretty well, and can be removed as easily as they could be originally: